Stroop color-word task
A task family contrasting color naming when written words agree or conflict.
Performance task The format describes how observations are obtained, not an evidence rating.
What it measures
Interference during selective responding, with both speed and accuracy relevant.
What people do
A participant reports ink color while a written color word may compete. Baseline choice, response mode and trial arrangement define the actual task.
What the scores mean
A latency or completion-time contrast estimates interference under that protocol. A difference score is not a direct reading of a single inhibition capacity.
This explains the logic of a score. It does not supply a test form, individual result or administration manual.
Development and versions
John Ridley Stroop’s 1935 experiments separated interference with reading from interference with color naming. Modern clinical editions and computerized variants are distinct.
Evidence: (Stroop (1935): Studies of Interference in Serial Verbal Reactions, n.d.); (Hedge et al., 2017); (Spielman et al., 2020)
Reliability and validity evidence
The original experiments document the interference phenomenon. Hedge and colleagues show why a reproducible group effect need not yield reliable individual rankings. Their conclusion concerns measured implementations, not every Stroop variant.
Evidence belongs to a version, sample and interpretation. Read the source notes for what was actually inspected.
Evidence: (Stroop (1935): Studies of Interference in Serial Verbal Reactions, n.d.); (Hedge et al., 2017); (Spielman et al., 2020)
Where it is used
Experimental studies of attention and conflict; standardized clinical versions require their own evidence and norms.
Limits and common misreadings
- Reading fluency, color vision, response speed and language can matter.
- Speed without error information can mislead.
- Do not convert an informal web demo into a clinical score.
Read an example
Original teaching example
An original synthetic comparison has mean response times of 500 and 580 milliseconds across two conditions. The 80-millisecond difference is a task contrast, not a percentile or diagnosis. The lab explains why individual reliability is a separate question.
Constructed for learning; this is not a result from the instrument’s validation study.
Access and permission
The historical article is accessible as a university transcription. Commercial task materials, scoring rules and norms can have separate rights.
Textbook connection
Related contextPsychology 2e: 2.2 Approaches to Research
Conceptual context; the instrument-specific sources in this profile provide the primary evidence.
Related measures
Read a measure critically · Explore score examples · Review research methods
Sources and inspection notes
Checked October 22, 2026. An inspection date is not a publication date. Publisher information is not presented as independent validation; abstract-only and indexed-excerpt access are identified below.
- Hedge, C., Powell, G., & Sumner, P. (2017). The reliability paradox: Why robust cognitive tasks do not produce reliable individual differences. Behavior Research Methods, 50(3), 1166–1186. https://doi.org/10.3758/s13428-017-0935-1Consulted sourcePrimary research: open article abstract and conceptual introduction
Distinguishes dependable experimental effects from reliable individual differences. Online publication was 2017; journal issue is 2018. No universal reliability coefficient assigned to task families.
- Spielman, R. M., Jenkins, W. J., & Lovett, M. D. (2020). Psychology 2e. OpenStax. https://openstax.org/books/psychology-2e/pages/2-2-approaches-to-researchTextbook: conceptual context
Conceptual context; the instrument-specific sources in this profile provide the primary evidence.
- Stroop (1935): Studies of interference in serial verbal reactions. (n.d.). https://www.yorku.ca/pclassic/Stroop/Primary article: university-hosted transcription
Original color-word interference procedure and distinction between reading and color naming inspected. Modern test editions require their own manuals.